VARIATION AND SELECTION 17 



large number of individuals has been examined 

 with regard to any particular organ, structure, or 

 character, the observed variations may form a con- 

 tinuous series, that is to say, practically every 

 intermediate degree of magnitude from the smallest 

 to the greatest that occurs may be found ; or the 

 series may be discontinuous, that is to say, there 

 may be large gaps between any two variations of 

 the series. I have been using the convenient term 

 magnitude, but the variations concern not only 

 characters that can be measured by weight and 

 length, but those distinguished by chemical differ- 

 ences revealing themselves in colour and function. 

 In the easy case of colour, the continuous series may 

 appear as a steady passage across the chromatic 

 scale, one tint passing imperceptibly into the next ; 

 the discontinuous series as a set of distinct colours. 

 Finally, the observed variations may be scattered 

 evenly round a central point, as if they were what we 

 should be disposed to call chance fluctuations, or 

 they may be oriented in a definite direction, as if 

 the constitution of the creature were obeying a 

 guiding impulse. A. R. Wallace believed that the 

 small, continuous, and unoriented variations were 

 the material of natural selection. Darwin differed 

 in his opinion at various times in his long career, but 

 on the whole was inclined to the view that Wallace 

 urged. Since the death of Darwin a very large 

 amount of work has been carried out in the analysis 

 of variations, and methods have been devised by 

 which it has been possible to conduct experimental 

 enquiry into the behaviour of characters in inherit- 

 ance on a scale and with an exactness that have 



